During the weekly meeting, someone clearly raises the issue: a key process has been stalled for three weeks and requires a decision.
Everyone nods. Everyone agrees that it is urgent. The meeting ends with a phrase such as, “We’ll follow up,” which sounds like an agreement.
One week later, the same problem returns to the table. No one resolved it, and no one clearly remembers why.
It is not that someone forgot. Somewhere between “we all agree” and concrete action, responsibility dissolved among the people in the room.
This is one of the most silent problems affecting teams: they do not fail because of a lack of talent, information, or agreement, but because no one ultimately becomes the owner of the decision.
A Phenomenon Supported by Decades of Research
The diffusion of responsibility was documented by psychologists John Darley and Bibb Latané in 1968, when they studied emergency situations in which the presence of multiple witnesses reduced the likelihood that any one person would act.
The more people believe they could intervene, the less individual responsibility each person may feel, because they assume someone else will take action.
In organizations, the context is different, but the pattern is recognizable:
- Everyone was informed.
- Several people participated.
- No one objected.
- No one was explicitly placed in charge.
A meeting, an email with twenty people copied, or a committee requiring unanimous agreement can become corporate versions of the same dynamic.
More Participants Do Not Always Lead to Better Decisions
Complex decisions require different perspectives. Finance, Sales, Operations, Technology, and Human Resources may all contribute essential information.
However, involving more people does not automatically improve the quality of the decision-making process.
Bain & Company proposes what it calls the “Rule of 7”: when a decision-making group exceeds seven people, each additional participant may reduce decision effectiveness by 10%.
The problem is not only the size of the group. It is also the ambiguity surrounding the role each participant is expected to play.
Bain’s RAPID model distinguishes five functions:
- Recommend: Develop and present a proposal.
- Agree: Validate mandatory requirements, such as legal or regulatory conditions.
- Provide Input: Contribute data, expertise, and information about potential impacts.
- Decide: Make the final decision and commit the organization.
- Perform: Implement what has been decided.
Bain recommends that, ideally, only one person should have the final decision-making authority. When this role belongs to a group, the method for resolving disagreement should be defined in advance.
A frequent mistake is to involve too many people in the consultation role while leaving the decision-maker’s role shared or unclear.
Each participant may honestly claim that they fulfilled their responsibility. Even so, the outcome remains without an owner.
Paralysis Is Also Produced by the System
Indecision is often attributed to insecure leaders or people who lack initiative. However, it frequently results from organizational conditions.
Ambiguous Authority
People understand their responsibilities, but they do not know the actual limits of their autonomy.
Conflicting Objectives
A decision may benefit the customer while negatively affecting one department’s performance indicator. Each function then protects its own results.
A Culture of Punishment
When making a mistake has more serious consequences than failing to act, postponing decisions becomes a rational behavior.
Constant Escalation
Employees are asked to act autonomously, but their decisions are reviewed or reversed whenever they do not align with the leader’s preferences.
Consensus as Protection
Listening to different perspectives is necessary. Obtaining everyone’s approval for every decision, however, can become a way of avoiding individual exposure.
The problem is not that there are no people willing to decide. The problem is that the system does not clearly establish who has both the right and the obligation to do so.
AI Adds a New Layer of Ambiguity
Artificial intelligence can compare scenarios, analyze information, and generate recommendations in a matter of minutes.
However, it introduces a critical question:
Who actually made the decision?
Madeleine Clare Elish uses the term moral crumple zone to describe a situation in which the person closest to an automated system ends up absorbing responsibility for an outcome over which they had limited control.
The tool makes a recommendation, other people design or configure it, the organization defines the conditions under which it will be used, and an employee approves the result. However, when something goes wrong, responsibility may fall entirely on the human operator.
The opposite may also occur: responsibility becomes so widely distributed among the tool, the provider, the analyst, the team, and the leader that no one considers themselves fully accountable.
Artificial intelligence does not eliminate the diffusion of responsibility. It can deepen it.
For this reason, before using an AI-generated recommendation, the team should clarify:
- Who validates the data and the result.
- What level of autonomy the tool has.
- Which human criteria cannot be delegated.
- Who makes the final decision.
- Who monitors its effects.
- How the process will be documented.
The OECD establishes that organizations and individuals that develop, deploy, or operate AI systems should be accountable according to their role, context, and ability to act. It also recommends ensuring the traceability of data, processes, and decisions.
Using AI does not reduce human responsibility. It requires responsibility to be distributed more precisely.
The Cost of Leaving the Problem Unresolved
When responsibility is diluted, decisions are not delayed occasionally. They are delayed systematically.
Problems return to meetings, projects wait for additional approvals, and opportunities lose value while the team searches for a level of certainty that will probably never arrive.
There is also a less visible cost: the loss of learning.
A team learns when it makes a decision, observes the consequences, and adjusts its criteria. If no one can explain who made the decision, what information was used, or why one alternative was selected, it is also impossible to understand what should be done differently.
Without clear responsibility, mistakes do not become learning opportunities.
They become another round of controls, meetings, and approvals.
Naming the Pattern Without Looking for Someone to Blame
The usual response is to ask one person to be more decisive or show greater initiative. This solution confuses a systemic pattern with an individual characteristic.
Before demanding accountability, the organization must design the conditions that allow people to exercise it:
- Classify decisions according to risk, impact, and reversibility.
- Define what information is sufficient to take action.
- Assign an owner to every decision.
- Limit consultation to those who provide genuine value.
- Establish how AI will be used within the process.
- Review outcomes without turning every mistake into blame.
- Prevent leaders from punishing the autonomy they claim to promote.
Accountability is not merely about identifying someone to blame. It involves being able to explain how a decision was made, which criteria were used, and what learning the outcome produced.
What Does Team Coaching Contribute?
Consulting can design governance structures, responsibility matrices, and policies for the use of AI.
Team coaching addresses another dimension: it helps the team observe what it does when it must make a decision.
It can explore questions such as:
- At what point does responsibility become diluted?
- Which decisions do we escalate out of necessity, and which do we escalate out of fear?
- What story supports the belief that “it is better to decide together”?
- How do we respond to disagreement?
- What happens when AI contradicts the team’s experience?
- Who should be responsible for the decision before the analysis begins?
The coach does not decide on behalf of the team or replace leadership authority. The coach helps team members identify the pattern, agree on responsibilities, and learn how to disagree, decide, and respond without looking for someone to blame.
From Shared Responsibility to Decisions with Clear Ownership
At Euro Business Coach, we understand that the ability to make decisions connects human capital with structural capital.
We work on processes, roles, criteria, and governance mechanisms, but also on the trust required so that taking ownership of a decision is not experienced as political exposure or personal risk.
Shared responsibility may appear collaborative. However, when no one has the final decision-making authority, the outcome is often left without an owner.
A team begins to move forward when it stops asking only:
“Do we all agree?”
And begins asking:
“Who, specifically, owns this decision?”
AI can expand the available alternatives and improve the analysis.
The responsibility to choose remains human.
